Effective approach to predict soil-water retention curve of bentonites considering adsorption and capillarity

Abstract

Understanding soil–water retention behavior is a longstanding topic. Water retained in soils can be decomposed into adsorptive and capillary components, controlled by different physicochemical mechanisms. The capillary water retention was frequently discussed in literatures, but the adsorption role was rarely considered, especially for high active clays (e.g., bentonite). In this study, two novel equations for quantifying adsorptive and capillary water retentions are proposed, generating a twofold model to continuously simulate soil–water retention behavior of bentonites. Only 5 parameters, i.e., the maximum adsorption capacity, characteristic adsorptive and capillary suctions, and uniformities of adsorptive and capillary pores, are defined with clear physical meanings. A graphical method was suggested to firstly determine the maximum adsorption capacity, and the remaining parameters are efficiently estimated by non-linear curve fitting. The water retention data for various bentonites, representing a variety of hydration conditions, initial compactness, montmorillonite content and suction range, are used to assess the model performance. The predictions agree well with the measured total, adsorptive and capillary water contents of a Wyoming bentonite, and the fitting curves also match well with test data for other bentonites over the full suction range. Adsorption parameters are distributed within a narrow range, while the characteristic capillary suction is also distributed within a limited range under constant volume condition. The proposed model allows reasonable predictions about the capillary onset and the transition from adsorption to capillarity, revealing obvious superiority by comparison with other hybrid models. This work offers a new pathway to quantitively assess the soil–water retention curve of high active clays.